Lithium and valproate are the most commonly used drugs for the treatment of bipolar disorder. The precise mechanisms underlying their clinical efficacy remain to be defined. We first investigated the neuroprotective effects of lithium against excitotoxicity elicited by glutamate, a major excitatory amino acid neurotransmitter involved in the synaptic plasticity and pathogenesis of neurodegenerative and neuropsychiatric disorders. We found that long-term exposure to lithium chloride dramatically protects cultured rat cerebellar granule cells (CGCs), against glutamate-induced excitotoxicity which involves apoptosis mediated by N-methyl-D-aspartate (NMDA) receptors. In CGCs, the lithium neuroprotection occurs at therapeutically relevant concentrations (0.5-5.0 mM) and requires treatment for 6-7 days for complete protection to occur. In CGCs, the lithium-induced neuroprotection against glutamate excitotoxicity involves multiple mechanisms. These include inactivation of NMDA receptors, induction of cytoprotective Bcl-2 but down-regulation of the pro-apoptotic proteins p53 and Bax. In addition, lithium activates the cell survival factor Akt and CREB through their enhanced phosphorylation. In contrast, glutamate treatment induces effects opposite to those induced by lithium. In addition, we found that glutamate induces a robust activation of AP-1 binding, Jun N-terminal kinase (JNK) and p38 kinase. Suppression of these glutamate-induced activities using selective inhibitors results in neuroprotection, suggesting their roles in excitotoxicity. Moreover, long-term lithium-induced neuroprotection is concurrent with inhibition of glutamate-induced activation of AP-1 binding, JNK and p38 kinase. Collectively, our results suggest that glutamate excitotoxicity involves induction of pro-apoptotic genes and suppression of cytoprotective gene expression. Furthermore, lithium neuroprotection is due, at least in part, to suppression of these glutamate-induced effects. Using SYM-2081, a glutamate uptake blocker and kainate receptor agonist, we found that valproate, another mood stabilizer, blocked SYM-2081-induced excitotoxicity in CGCs. Moreover, valproate-induced neuroprotection is mimicked by inhibitors of histone deacetylase (HDAC), suggesting the effects are mediated through inhibition of HDAC. Additionally, in CGCs we found that under conditions in which neither lithium nor valproate alone is effective in protecting against glutamate excitotoxicity, combined treatment with lithium and valproate provides a synergy in neuroprotection. A recent study showed that VPA induces a-synuclein in CGCs and this induction has a neuroprotective role against glutamate excitotoxicity. Moreover, the VPA-induced a-synuclein expression is mimicked by other HDAC inhibitors such as butyrate and trichostatin A, and involves an increase in histone H3 acetylation in the a-synuclein promoter and a robust activation of a-synuclein promoter activity. We have extended our studies to include primary cultures of cortical neurons prepared from embryonic rats. These cortical neuronal cultures are highly vulnerable to glutamate-induced apoptotic insults. Pretreatment with subtherapeutic or therapeutic concentrations of LiCl for 6 days robustly protects against glutamate excitotoxicity in these cultures. Thus, significant protection was achieved at 0.1-0.2 mM with a nearly complete protection at 1.0 mM. The lithium neuroprotection in cortical neurons is associated with a reduction in NMDA receptor-mediated calcium influx, and this lithium-induced action is correlated with a selective decrease in tyrosine phosphorylation at position 1472 of the NR2B subunit of the receptor. The latter is preceded by a reduction in the activation of Src which is involved in NR2B subunit tyrosine phosphorylation. Our results suggest that modulation of glutamate receptor hyperactivity represents, at least in part, the molecular mechanisms by which lithium alters brain function and exerts its clinical efficacy in the treatment for manic depressive illness. These novel actions of lithium also suggest that excessive glutamatergic neurotransmission may be the pathogenic mechanism underlying bipolar illness. The neuroprotective effect of lithium requires the expression of brain-derived neurotrophic factor (BDNF) and activation its receptor TrkB. The induction of BDNF is preceded by an increase in BDNF exon III mRNA and its transcriptional promoter activity. Using rat cortical cell and CGC cultures, we also demonstrated that lithium enhances the proliferation of their progenitor cells. Furthermore, the decrease in neuronal progenitor proliferation induced by glutamate, glucocorticoids or haloperidol is antagonized by lithium pretreatment. The neuroprotective effects have also been shown with valproate, another mood stabilizer. We found that valproate robustly protects against glutamate excitotoxicity and spontaneous cell death in cultured brain neurons. Additionally, valproate induces BDNF through exon III activation. In a collaborative study with Dr. Jau-Shyong Hong?s group in NIEHS, NIH, we demonstrated that valproate protects midbrain dopaminergic neurons from LPS-induced inflammation mediated through microglia activation. Thus, valproate has anti-inflammatory effects in the cellular model of Parkinson?s disease. The above in vitro studies have been extended to studies using animal models of neurodegenerative diseases. We investigated the neuroprotective effects of mood stabilizers in a rat model of stroke which is the third leading cause of death in the US. We performed middle cerebral artery occlusion (MCAO) in rats, a procedure which triggers brain infarction and results in neurological deficits. Our results showed that pre- or post-treatment with lithium robustly reduces MCAO- induced infarct volume and suppresses the neurological deficits detected in motor, sensory and reflex texts. The lithium-induced neuroprotection is associated with induction of heat shock protein 70 (HSP70) and inhibition of caspase-3. A more recent study showed that valproate has similar neuroprotective actions in the rat MCAO model of stroke: it reduces both the infarct volume and neurological deficit scores, and induces HSP70. The beneficial time window for both drugs is at last three hours after the onset of ischemia. The valproate-induced neuroprotection against ischemic insult is mimicked by another HDAC inhibitor, sodium butyrate, again suggesting a neuroprotective role of HDAC inhibition. Importantly, in our MCAO studies, we confirmed the anti-inflammatory effects of valproate as revealed by suppression of ischemia-induced microglia activation. In the Huntington's disease animal model, we injected quinolinic acid, a partial agonist of the NMDA receptor, into the left side of rat striatum. This quinolinic acid-induced lesion requires activation of the transcription factor NF-kB and induction of p53, c-Myc and cyclin D1 and is protected by metabotropic receptor agonists and prostaglandin A1. Our results show that pretreatment with lithium for 16 days or one day decreases the size of striatal lesion by 40-50%. In addition, lithium neuroprotection effects are associated with over-expression of striatal Bcl-2. The neuroprotection is also correlated with suppression of QA-induced DNA damage and caspase-3 activation. Additionally, lithium induces enhanced cell proliferation in the striatum near the site of quinolinic acid injection. Thus, our in vitro and in vivo studies raise the possibility that lithium, in addition to its use for bipolar disorder, may have expanded use for the treatment of neurodegenerative diseases, particularly those linked to excitotoxicity, such as stroke and Huntington?s disease.